Proton therapy is considered, to a large extent, to be the radiation therapy of the choice for treating pediatric brain tumors. This is mostly because its dose confinement to the target is better than that of the high energy x rays due to its Bragg peak feature, although the slightly higher relative biological effectiveness (RBE) of protons at the target is also helpful. However, these advantages still do not make proton therapy an ideal method of pediatric brain tumor therapy because of the fact that children undergoing brain tumor therapy with protons still exhibit a certain level of late cognitive deficits. One explanation coul be that protons lack the shallow-tissue-sparing effect that the MV x rays have. The method proposed here, namely proton therapy with minibeams, turns the solid incident beams of protons into arrays of parallel, thin (0.3 mm) planes of protons called minibeams that spare shallow tissues. As these minibeams penetrate the tissues they gradually broaden because of multiple Coulomb Scattering, and depending on the spacing between them they merge with each other at a certain depth. This depth can be adjusted to be proximal to the target's location. Because the children's cognitive damage comes to a large extent from radiation damage to their cortex, and because proton minibeams should spare the cortex to a large extent, it is expected that proton minibeam therapy will reduce cognitive deficits in children. The method is completely compatible with Bragg-peak spreading. The optimal beam spacing between the 0.3-mm-thick proton minibeams in the arrays is between 0.7 and 1.0 mm. The 0.7 mm spacing value is chosen because it is known from studies with synchrotron minibeams that the minibeam's tissue sparing effect will start to decline beyond 0.7 mm minibeam thickness, and therefore 0.7-mm beam spacing on-center will produce the largest proton collimator yield without compromising the array's shallow-tissue sparing. We propose to test this technique in animal models of brain tumors and/or normal brain based on the hypothesis that sparing the cortex while treating deep-seated brain tumors might reduce the neurocognitive toxicities of treatment. In Aim 1 we compare the radiation effects between proton minibeams and solid beams on the entire brains of the radiosensitive transgenic mice APOE4. In Aim 2 we will treat the intracranial malignant rat brain tumor 9L gliosarcoma comparatively with proton minibeams, merging on the proximal side of the target, and proton solid beams. The head of a normal rat will be positioned proximal to that of the tumor-bearing rats. Collectively these experiments should evaluate the method's potential for clinical use. Successful completion of the proposed research could give a new dimension to proton therapy and could have broad clinical applications for proton therapy ranging from the treatment of pediatric brain tumors to hypofractionated regimens for a variety of tumors whose shallow frontal tissues can be immobilized.